SYSTEMS OF MATING 55 



probability of incrosses would be approached more readily. If there were some selection 

 for heterozygotes, incrosses would be obtained less readily, but could be achieved with 

 brother-sister inbreeding. We know, in other words, what would be the direction of 

 the change when these other factors are taken into consideration, but we have not 

 worked out the probabilities in detail. 



Dr. Crow: It is relevant to point out in this connection that, if an organism has 

 something like 10,000 genes, one cannot apply a strong selection to each of them. 

 If the effects of inbreeding on decreasing heterozygosity of any particular locus are, say, 

 25 per cent per generation, this means that even intense selection cannot counteract 

 the increased homozygosity of more than a small fraction of these in any one population. 



Dr. Wright: It should be emphasized that these methods have nothing to do with 

 lethals or semilethals. Dr. Slatis has worked out a different coefficient dealing with 

 lethals in bison. The average rate of fixation of loci as derived by these methods 

 applies strictly only to neutral alleles but applies approximately when one allele has 

 a selective advantage of lower order than the reciprocal of ten times the effective size 

 of the inbreeding group. This includes most of the loci, under close inbreeding, for 

 reasons indicated by Dr. Crow. With stronger selection, the rate for the more 

 favorable allele increases while that for the unfavorable one decreases, with only 

 slight change in the total, unless the selection is rather severe. With strong selection 

 for the heterozygote over both homozygotes, the total rate is, of course, decidedly 

 lowered just as it is under enforced heterozygosis of a strongly linked gene, the 

 problem investigated by Green and Doolittle. 



Dr. Gowen: This approach is really based on an infinite population of gametes, 

 breeding entirely at random. From an evolutionary point of view perhaps the assump- 

 tions may be right, but actual breeding experiments with guinea pigs, mice, Drosophila, 

 and so forth never meet the conditions which have been postulated. Selection continu- 

 ally occurs, however random it may be, with the consequence that individually, 

 inbreeding coefficients may be quite misleading. 





